A new pliopithecid genus (primates: pliopithecoidea) from castell de barberà (vallès-penedès basin, catalonia, spain)

Abstract

Barberapithecus huerzeleri gen. et sp. nov. (Primates, Pliopithecidae) is erected on the basis of material from Castell de Barberà (Middle to Late Miocene, ca. 11.2–10.5 Ma), in the Vallès-Penedès Basin (Catalonia, Spain), including: 15 teeth (representing most of the permanent dentition) from a single female individual (holotype); an isolated P/3 (paratype); and a male C1/ (referred to the hypodigm). Previously, this material had been only partially figured and described, being attributed to Pliopithecus or to a new taxon with possible crouzeliine affinities. The erection of a new genus is justified by several autapomorphic features, such as markedly buccolingually compressed and mesiodistally elongated C1/, extremely buccolingually compressed, and mesiodistally oriented C/1 main cusp, P/4 with a large trigonid subequal to the talonid, very large distal foveae on the M/1 and especially the M/2, and lower molars with a quadrangular central fovea and a mesially situated entoconid. These features are associated with a set of crouzeliine synapomorphies, such as buccolingually compressed and peripheralized cusps, well-developed crests, large and well-defined occlusal foveae, upper molars with long preprotocrista, short hypoparacrista, somewhat distally situated protocone and short distal fovea, distinct P/3 metaconid, well-developed P/4 premetacristid, and relatively narrow lower molars with a reduced entoconid. Although more primitive, Barberapithecus resembles Anapithecus in some derived features. Both taxa are included into a new tribe (Anapithecini), together with other crouzeliines except Plesiopliopithecus (tribe Crouzeliini). The retention of primitive, pliopithecine-like features in Barberapithecus suggests that anapithecins might have evolved from a Pliopithecus ancestor, so that as currently conceived the Crouzeliinae might be polyphyletic. Am J Phys Anthropol 2012. © 2011 Wiley Periodicals, Inc.

Pliopithecid dental remains from the Miocene site of Castell de Barberà (CB; Barberà del Vallès, Catalonia, Spain) were first reported more than thirty years ago (Anonymous,1975a,b; Crusafont-Pairó,1975,1978). Isolated teeth from CB, initially corresponding to a single individual, were partially described, together with an isolated dP/4 from the similarly aged locality of Can Feliu (Sant Quirze, Catalonia, Spain), by Crusafont-Pairó and Golpe-Posse (1981,1982). These authors described the lower dentition, noting similarities with Pliopithecus lockeri (now Plesiopliopithecus lockeri), although merely assigning the remains to Pliopithecus sp. (which currently would mean Pliopithecidae indet., following the more restricted current usage of the genus Pliopithecus). In fact, these authors suggested that these remains likely represented a new form among European pliopithecids, but refrained from formally erecting a new species. Crusafont-Pairó and Golpe-Posse (1981) even mentioned an ongoing study of the upper dentition that would clarify the taxonomic status of this material, but this study was never published. Later, Ginsburg (1986) attributed these specimens to Crouzeliinae indet. nov., while they were referred to Pliopithecus aff. antiquus by Moyà Solà et al. (1990), to P. antiquus by Andrews et al. (1996), to Pliopithecinae gen. et sp. indet. (new) by Begun (2002), to P. cf. antiquus by Harrison et al. (2002: footnote 1), and most recently to Pliopithecidae nov. by Alba et al. (2010). In this article, we describe and figure in detail both the upper and lower dentition of the pliopithecid from CB, and on this basis, we erect a new genus and species. A new tribe is further erected to include the new genus and all previously known crouzeliines, with the exception of Plesiopliopithecus (see later).

AGE OF CASTELL de BARBERÀ

There are serious uncertainties regarding the age of CB, which cannot definitively be settled at that moment, given the lack of magnetostratigraphic data for this locality. Since its discovery, CB has been considered stratigraphically intermediate between Sant Quirze (MN8, late Aragonian, Middle Miocene) and Can Llobateres 1 (CLL1; MN9, early Vallesian, Late Miocene; e.g., Crusafont-Pairó and Golpe,1972). Given the purported lack of hipparionin equids at this locality (but see later), Crusafont Pairó and Truyols Santonja (1951) attributed CB to the late Aragonian (=“Vindobonian”; see also Crusafont-Pairó and Golpe,1972; Crusafont, 1985). An attribution to the latest Aragonian (MN8 sensu Mein and Ginsburg,2002), has been favored by most subsequent authors (e.g., Agustí et al.,1985,1997,2001; Moyà-Solà and Agustí,1987, 1990; Moyà-Solà et al.,1990; Begun,2002; Alba et al.,2006,2010,2011b; Casanovas-Vilar,2007; Casanovas-Vilar et al.,2011), although CB has been alternatively assigned to the MN9 by some (De Bruijn et al.,1992; Andrews et al.,1996), mainly as a result of the faunal similarities between the rodent assemblages of CB and CLL1 (Aguilar et al.,1979). In fact, the relative stratigraphic proximity between CB and CLL1 (110 m according to Crusafont Pairó and Truyols Santonja, 1951: Fig. 5; only 30 m according to Crusafont-Pairó and Golpe,1972) more strongly argues against a late Aragonian age for the former, unless the existence of an undocumented large fault is hypothesized. This is because CLL1 corresponds to the early Vallesian (the local biozone MN9b of Agustí et al.,1997, characterized by the presence of Cricetulodon), with an estimated age of 9.7 Ma on the basis of magnetostratigraphic correlation (Agustí etal.,1996,1997; Casanovas-Vilar et al.,2011; Alba et al.,2011a). The entry of Hippotherium in the Vallès-Penedès basin, which signals the beginning of the Vallesian (Late Miocene), is dated at 11.1 Ma (Garcés et al.,1996; Agustí etal.,1997,2001; Casanovas-Vilar et al.,2006), thereby implying an age gap of 1.4 myrs between both sites if CB is considered to be Aragonian in age. This figure is almost thrice the age difference of 0.55 myrs that can be computed on the basis of a 110 m stratigraphic separation and the average sedimentation rate of 0.2 m/kyr computed by Garcés et al. (1996) for the Vallès-Penedès Basin, making it unlikely a late Aragonian (MN8) age for CB. On this basis alone, CB might be as young as 10.3 Ma, although the lack of Cricetulodon conclusively indicates an age older than 10.4 (i.e., MN9a; Garcés et al.,1996; Agustí et al.,1997).

Even a well-substantiated lack of Hippotherium (Hipparion s.l.) is not necessarily indicative of an Aragonian age, given the lack of Hippotherium remains from the earliest Vallesian levels from the Calatayud-Daroca Basin, where this taxon is not recorded until about 10.8–10.7 Ma (Garcés et al.,2003), as well as in the early Vallesian (ca. 11.0–10.5 Ma) local stratigraphic series of Ecoparc de Can Mata (ECM) within the Vallès-Penedès Basin (Carmona et al.,2011). The ECM series is stratigraphically situated above the Aragonian series of Abocador de Can Mata and the classical localities Can Mata I and III, between which the Aragonian/Vallesian transition is comprised (Alba et al.,2009; Moyà-Solà et al.,2009), thus illustrating the rarity of this equid during the earliest Vallesian. Indeed, the age of CB can be further limited by the find of an eroded, surface-collected fragmentary fossil of Hipparion from CB (Crusafont-Pairó and Golpe-Posse,1974; Rotgers and Alba,2011). According to Crusafont-Pairó and Golpe-Posse (1974), the Hipparion remain did not come from the main fossiliferous layer, situated at about the middle of the 20m-thick stratigraphic section of CB, but from a somewhat higher layer within the series. These authors hypothesized that either this was the earliest (Aragonian) Hipparion record in the basin, or the Aragonian/Vallesian transition was situated along the series between both fossiliferous layers. They did not take into account the obvious alternative: that the lower layer from CB, from which the pliopithecid teeth described here come from, was already Vallesian, even though Hippotherium was not recorded there. Crusafont-Pairó and Golpe-Posse (1974) implicitly considered this to be highly unlikely, given that no hipparionin remains had been found among thousands of fossils recovered from the main fossiliferous layer. Nevertheless, the fact is that Hippotherium seems to be remarkably rare during the earliest Vallesian (the local biozone MN9a of Agustí et al.,1997), only becoming much more frequent during the late early and the late Vallesian (MN9b and MN10, respectively). Although Crusafont-Pairó and Golpe-Posse (1974) did not figure the hipparionin fragmentary remain from CB, a Hippotherium molar from this locality was recently located among the collections from the Institut Català de Paleontologia Miquel Crusafont, ICP, (Rogers and Alba,2011). As such, the main fossiliferous level of CB must be very close to theearly Vallesian level (if not Vallesian itself), which is stratigraphically situated only a few meters above. Only the record at CB of the cervid Euprox aff. furcatus (Azanza and Menéndez,1990), instead of the typically Vallesian E. dicranocerus (Moyà-Solà and Agustí,1987; Azanza and Menéndez,1990), is suggestive of a late Aragonian instead of early Vallesian age for this locality. Nevertheless, this is not a reliable criterion, because E. furcatus, widespread during the Middle Miocene, further survived until the early Late Miocene, coexisting in time with E. dicranocerus (Vislobokova,2007).

Micromammals are not of much help for resolving this biostratigraphic issue. On the basis of the lack of Hippotherium and Cricetulodon, the rodent assemblage from CB would be attributed to the Megacricetodon ibericus + Democricetodon crusafonti local biozone (Alba et al.,2006; Casanovas-Vilar,2007; Casanovas-Vilar et al.,2011), which is correlated to the MN8 sensu Mein and Ginsburg (2002; i.e., based on La Grive-Saint-Alban L3 as the reference fauna). However, it should be taken into account that latest Aragonian (MN8) and earliest Vallesian (MN9a) rodent faunas are very similar except for the presence of hipparionins in the latter (Agustí et al.,1997,2001; Casanovas-Vilar et al.,2006). The dispersal of this equid through the Old World was not accompanied by a significant faunal turnover of either micromammals or other macromammals (Agustí et al.,1997,2001), so that there is no objective basis to distinguish between MN8 and early MN9 localities from the Vallès-Penedès Basin based on micromammals alone. Given the above-mentioned caveats regarding the presence/absence of Hippotherium at the main fossiliferous level of CB, we provisionally favor an estimated age range of 11.2–10.5 Ma (MN8 or MN9a, latest Middle or earliest Late Miocene) for CB, as constrained by the entry of Hippotherium at 11.1 Ma and that of Cricetulodon at 10.4 Ma. Paleomagnetic sampling of the CB series would be required in the future, to constrain further the estimated age of the pliopithecid remains described in this article.

SYSTEMATIC PALEONTOLOGY

• Order PRIMATES Linnaeus,1758

• Suborder ANTHROPOIDEA Mivart,1864

• Infraorder CATARRHINI É. Geoffroy Saint-Hilaire,1812

• Superfamily PLIOPITHECOIDEA Zapfe,1961

• Family PLIOPITHECIDAE Zapfe,1961

• Subfamily CROUZELIINAE Ginsburg and Mein,1980

• Tribe ANAPITHECINI trib. nov.

• Type genus: Anapithecus Kretzoi,1975.

• Other included genera: Laccopithecus Wu and Pan,1984, Egarapithecus Moyà-Solà et al.,2001, and Barberapithecus gen. nov.

Diagnosis

Crouzeliine subtribe that differs from the nominotypical one, Crouzeliini (including the genus Plesiopliopithecus Zapfe,1961, of which Crouzelia, Ginsburg,1975, is considered a junior subjective synonym) by the presence of a distinct P/3 metaconid, the deep trigonid foveae on the P/4 and lower molars, the molars with a reduced entoconid, and the retention of several plesiomorphic features such as the less reduced cingulids (except in Laccopithecus).

BARBERAPITHECUS GEN. NOV.

Type species

Barberapithecus huerzeleri gen. et sp. nov.

Diagnosis

Small-sized pliopithecid species, similar in dental size to Dionysopithecus and several Pliopithecus species (e.g., P. canmatensis). The dentition shows buccolingually compressed cusps, well-developed crests, large and well-defined occlusal foveae, peripheralized cusps, and well-developed cingula in both the upper and lower molars. C1/ with a strongly buccolingually compressed and mesiodistally elongated main cusp, and an extensive lingual shelf. Upper molars with peripheralized protocone and hypocone, somewhat distally situated protocone, long preprotocrista, short hypoparacrista, mesially curved hypoparacrista, and short distal fovea. Extremely buccolingually compressed and mesiodistally oriented C/1. Distinct metaconid at the end of the P/3 hypoprotocristid. P/4 with a large trigonid (subequal to the talonid) and a well-developed premetacristid that lingually closes the trigonid. P/4 and lower molars with deep trigonid foveae. Lower molars with a distinct pliopithecid triangle (progressively more well-developed from M/1 to M/3), quadrangular central fovea, mesially situated and reduced entoconid, distally situated hypoconid (approximately in front of the entoconid on the M/2 and M/3), short posthypocristid, and large and centrally placed hypoconulid. Metaconid and protoconid more closely situated than the entoconid and hypoconid on the M/1, so that the trigonid basin is narrower than the central fovea, whereas on the M/2 and M/3 the metaconid and protoconid are wide apart (approximately like the entoconid and hypoconid). M/1 and especially M/2 with very large distal foveae. Distance between the protoconid and metaconid apices is only slightly smaller than that between those of the hypoconid and entoconid. M/3 with a reduced distal lobe.

Differential diagnosis

Barberapithecus differs from dionysopithecines by the following features: more peripheralized and buccolingually compressed cusps of the lower molars and lingual cusps of the upper molars; upper molars with a longer preprotocrista, a shorter hypoparacrista, a shorter distal fovea, and reduced cingula and enamel wrinkling; lingual cingulum of the upper molars ending at the apex of the hypocone instead of surrounding its base; distance between the apices of the protoconid and metaconid only slightly smaller than between those of the hypoconid and entoconid on the M/2; wider and generally larger mesial foveae on the M/2 and M/3; and M/3 absolutely longer and relatively narrower than the M/2. Barberapithecus differs from pliopithecines by the following features: more peripheralized and more buccolingually compressed cups in the upper and lower molars; more buccolingually compressed and mesiodistally elongated main cusp of the C1/; elongated and relatively narrower upper molars, with more peripheralized protocone and hypocone, a longer preprotocrista, a shorter hypoparacrista, and reduced cingula and enamel wrinkling; presence of a distinct P/3 metaconid; well-developed P/4 premetacristid that lingually closes the trigonid; relatively narrower lower molars, with longer and more extensive mesial foveae, sharper cristids, narrower buccal cingulids (particularly between the protoconid and hypoconid), shorter posthypocristid, and more mesially situated entoconid and more distally situated hypoconid (with both cusps aligned with one another, in particular on the M/2 and M/3). Finally, when compared to other crouzeliines, Barberapithecus can be distinguished by the less extremely buccolingually compressed cusps and crests, the slightly broader lower molars, the better-developed cingula in upper and lower molars, the more buccolingually compressed and mesiodistally elongated C1/ and C/1 main cusp, and the centrally situated hypoconulid on the lower molars (except from Plesiopliopithecus lockeri). More specifically, it can be distinguished from crouzeliins (Plesiopliopithecus) by several occlusal details, such as the much more developed cingula and the presence of a distinct P/3 metaconid, the deep trigonid fovea on the P/4 and the lower molars, and the reduced entoconid on the lower molars. With regard to the other anapithecin genera, Barberapithecus can be distinguished from them all by the possession of several autapomorphic features enumerated in the Diagnosis, as well as several additional occlusal details (see Results), most closely resembling the genus Anapithecus. Nevertheless, Barberapithecus can be further distinguished from the latter by additional dental features, such as the higher and narrower I1/, the higher-crowned C1/ with a more developed mesial sulcus, the more distally situated hypocone and better-developed lingual cingulum in the upper molars, the less elongated P/4 talonid, the more mesially situated entoconid and larger distal fovea in the lower molars (particularly the M/2), and the less elongated crown and central fovea as well as more developed buccal cingulid in the M/3.

Etymology

Genus name derived from “Barberà,” taken from the Catalan toponyms “Castell de Barberà” and “Barberà del Vallès”—which are respectively the type locality of the new species and the municipality where the this locality is situated—and from the Greek pithekos (ape).

Barberapithecus huerzeleri gen. et sp. nov.

(Figures 1, 2A–D, 3–11, 12F, and 13A–C).

Figure 1. Upper central incisor and female canine of Barberapithecus huerzeleri gen. et sp. nov. (holotype): (A–D) Right I1/ (IPS1724a), in mesial (A), labial (B), distal (C) and lingual (D) views; (E–H) Left C1/ (IPS1724b), in distal (E), buccal (F), mesial (G) and lingual (H) views. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Figure 2. Right male C1/ (IPS1823) tentatively included into the hypodigm of Barberapithecus huerzeleri gen. et sp. nov. (A–D), in distal (A), labial (B), mesial (C), and lingual (D) views, as compared to a right female C1/ (IPS1774) of Hispanopithecus laietanus from Can Llobateres 1 (E–H), in distal (E), labial (F), mesial (G), and lingual (H) views. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Figure 3. Lower canines of Barberapithecus huerzeleri gen. et sp. nov. (holotype): (A–D) Left C/1 (IPS1724i), in mesial (A), buccal (B), distal (C), and lingual (D) views; (E–H) Right C/1 (IPS1724h), in mesial (E), buccal (F), distal (G), and lingual (H) views. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Figure 4. Cheek teeth of Barberapithecus huerzeleri gen. et sp. nov. (holotype and paratype) in occlusal view: (A) Right P/3 (IPS34548); (B) Left P/3 (IPS1724k); (C) Right P/3 (IPS1724j); (D) Left P/4 (IPS1724l); (E) Left M/1 (IPS1724m); (F) Right M/2 (IPS1724n); (G) Left M/3 (IPS1724o); (H) Right M1/ (IPS1724c); (I) Right M2/ (IPS1724d); (J) Right M3/ (IPS1724f); (K) Left M2/ (IPS1724e); (L) Left M3/ (IPS1724g). All specimens are depicted with the mesial side towards the top. [Color figure can be viewed in the online issue, which is available at wileyonline library.com.]

Figure 5. Composite molar series of Barberapithecus huerzeleri gen. et sp. nov. (holotype): (A–C) Right upper series, in occlusal (A), buccal (B), and lingual (C) views; (D–F) Left lower series, in occlusal (D), buccal (E), and lingual (F) views. The M/2 corresponds to the mirror image of the right side, since the left M/2 is not preserved. The specimens are oriented with the mesial side towards the right in (A), (B), and (F), and towards the left in (C), (D), and (E). [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Figure 6. Drawings of the upper central incisor and canine of Barberapithecus huerzeleri gen. et sp. nov. (holotype): (A–E) Right I1/ (IPS1724a), in occlusal (A), mesial (B), labial (C), distal (D) and lingual (E) views; (F–J) Left C1/ (IPS1724b), in occlusal (F), mesial (G), buccal (H), distal (I), and lingual (J) views. Original artwork by Marta Palmero.

Figure 7. Drawings of the upper male canine referred to the hypodigm of Barberapithecus huerzeleri gen. et sp. nov. (IPS1823), in distal (A), labial (B), mesial (C), lingual (D), and occlusal (E) views. The discontinuous trace indicates an area where enamel is corroded, whereas the parallel lines in occlusal view depict a wear facet. Original artwork by Marta Palmero.

Figure 8. Drawings of the lower female canines of Barberapithecus huerzeleri gen. et sp. nov. (holotype): (A–E) Left C/1 (IPS1724i), in mesial (A), buccal (B), distal (C), lingual (D), and occlusal (E) views; (F–J) Right C/1 (IPS1724h), in mesial (F), buccal (G), distal (H), lingual (I), and occlusal (J) views. Original artwork by Marta Palmero.

Figure 9. Drawings of the lower premolars of Barberapithecus huerzeleri gen. et sp. nov. (holotype): (A–E) Left P/3 (IPS1724k), in occlusal (A), mesial (B), buccal (C), distal (D), and lingual (E) views; and (F–J) Right P/3 (IPS1724j), in occlusal (F), mesial (G), buccal (H), distal (I), and lingual (J) views; (K and L) Left P/4 (IPS1724l), in occlusal (K) and buccal (L) views. Original artwork by Marta Palmero.

Figure 10. Drawings of the lower molars of Barberapithecus huerzeleri gen. et sp. nov. (holotype): (A–C) Left M/1 (IPS1724m), in occlusal (A), buccal (B), and lingual (C) views; (D–F) Right M/2 (IPS1724n), in occlusal (D), buccal (E), and lingual (F) views; (G–I) Left M/3 (IPS1724o), in occlusal (G), buccal (H), and lingual (I) views. Original artwork by Marta Palmero.

Figure 11. Drawings of the upper molars of Barberapithecus huerzeleri gen. et sp. nov. (holotype): (A and B) Right M1/ (IPS1724c), in occlusal (A) and buccal (B) views; (C and D) Right M2/ (IPS1724d), in occlusal (C) and buccal (D) views; (E and F) Left M2/ (IPS1724e), in occlusal (E) and buccal (F) views; (G and H) Right M3/ (IPS1724f), in occlusal (G) and buccal (H) views; (I and J) Left M3/ (IPS1724g), in occlusal (I) and buccal (J) views. Original artwork by Marta Palmero.

Figure 12. Drawings of the lower postcanine series of several pliopithecines and crouzeliines: (A) Pliopithecus antiquus male (?) right (reversed) M/2-M/3 from La Grive PB A; (B) P. antiquus male left P/3-M/3 from Sansan (holotype); (C) P. canmatensis male right (reversed) P/3-M/3 (IPS35036) from ACM/C4-A1 (holotype); (D) P. canmatensis female left P/3-M/3 (IPS41719) from ACM/C4-Cb (paratype); (E) P. canmatensis male left P/3-M/3 (IPS41981) from ACM/C5-C3 (paratype); (F) Barberapithecus huerzeleri gen. et sp. nov., composite left lower postcanine series of the holotype, including: IPS1724j (reversed), IPS1724l, IPS1724m, IPS1724n (reversed) and IPS1724o; (G) Plesiopliopithecus auscitanensis from Sansan (holotype) left P/3-M/1. Original artwork by Marta Palmero; drawings (A) to (E) reproduced from Alba et al. (2010). All specimens drawn from originals, except by (A), (B) and (G), which were drawn from good quality casts. All specimens are depicted with the mesial side towards the top.

Figure 13. Environmental scanning electron microscope micrographs of molars of Barberapithecus huerzeleri gen. et sp. nov. (A–C, holotype) and Pliopithecus canmatensis (D, holotype), in occlusal view: (A) right M1/ (IPS1724c); (B) left M/1 (IPS1724m); (C) right M/2 (IPS1724n, reversed); (D) left M/2 (IPS35036). All specimens are depicted with the mesial side towards the top.

Diagnosis

As for genus.

Holotype

Fifteen teeth from a single female individual (IPS1724), including the right I1/ (IPS1724a; Figs. 1A–D and 6A–E), the left C1/ (IPS1724b; Figs. 1E–H and 6G–J), the right M1/ (IPS1724c; Figs. 4H, 5A–C, 11A–B, and 13A), the right (IPS1724d; Figs. 4I, 5A–C, and 11C–D) and left (IPS1724e; Fig. 4K and 11E–F) M2/, the right (IPS1724f; Figs. 4J, 5A–C, and 11G,H) and left (IPS1724g; Figs. 4L, and 11I,J) M3/, the right (IPS1724h; Figs. 3A–D and 8A–E) and left (IPS1724i; Figs. 3E–H and 8F–J) C/1, the right (IPS1724j; Figs. 4C and 9F–J) and left (IPS1724k; Figs. 4B and 9A–E) P/3, the left P/4 (IPS1724l; Figs. 4D and 9K,L), the left M/1(IPS1724m; Figs. 4E, 6D–F, 10A–C, and 13B), the right M/2 (IPS1724n; Figs. 4F, 5D–F, 10D–F, and 13C), and the left M/3 (IPS1724o; Figs. 4G, 5D–F, and 10G–I). See Table 1 for measurements. These teeth were partially figured and described by Crusafont-Pairó and Golpe-Posse (1981), and are currently housed at the Institut Català de Paleontologia Miquel Crusafont (ICP; Catalonia, Spain).

Table 1. Dental measurements (in mm) of the holotype (IPS1724), paratype (IPS1823), and other referred material (IPS34548) of Barberapithecus huerzeleri gen. et sp. nov. from Castell de Barberà (Catalonia, Spain)

Record No.	Tooth	MD	BL (m)	BLd	B/L (%)
B/L, BL/MD × 100; BL, buccolingual breadth; d, distal; l, left; m, mesial; MD, mesiodistal length; r, right. Values between parentheses are estimates due to tooth damage.
IPS1724a	r I1/	4.5	3.7	–	82
IPS1724b	l C1/	5.7	4.7	–	82
IPS1724c	r M1/	5.3	6.5	–	123
IPS1724d	r M2/	5.6	6.9	–	123
IPS1724e	l M2/	5.6	7.1	–	127
IPS1724f	r M3/	5.5	7.0	–	127
IPS1724g	l M3/	5.4	7.0	–	130
IPS1724h	r C/1	4.9	3.1	–	63
IPS1724i	l C/1	4.9	3.2	–	65
IPS1724j	r P/3	5.2	3.8	–	73
IPS1724k	l P/3	5.3	3.7	–	70
IPS1724l	l P/4	(5.0)	(4.1)	–	(82)
IPS1724m	l M/1	5.7	4.6	4.6	81
IPS1724n	r M/2	6.4	5.1	4.9	80
IPS1724o	l M/3	7.1	5.0	4.6	70
IPS1823	r C1/	8.1	6.2	–	77
IPS34548	r P/3	5.2	3.4	–	65

Type locality

Castell de Barberà (Barberà del Vallès, el Vallès Occidental, Catalonia, Spain), coordinates UTM 1 × 1 km 31T DF2897, placed in the hill of the Ripoll river, just below the demolished farmhouse of Ca n'Altimira (Crusafont,1975), and about 600–700 m NW of the Castell de Barberà, which gives its name to the locality.

Paratype

A right P/3 (IPS34548; Fig. 3A) from the type locality, also housed at the ICP, necessarily belongs to a different individual than the holotype. See Table 1 for measurements.

Other referred specimens

A right male C1/ (IPS1823; Figs. 2A–D and 7A–E), similarly housed at the ICP, is further referred to the hypodigm of the newly described taxon. Given the alternative interpretation of this specimen as a female hominoid C1/ (see Discussion), it is not formally designated as a paratype. Its measurements are reported in Table 1.

Distribution

Only recorded from the type locality.

Stratigraphic range

Age attribution of the type locality comprised between about 11.2 and 10.5 Ma, MN8 (late Aragonian) or MN9a (early Vallesian), close to the Middle to Late Miocene transition.

Etymology

Species epithet dedicated to the late Swiss paleontologist Johannes Hürzeler, in recognition of his great contribution to paleoprimatology.

DESCRIPTION

Upper dentition

Upper incisors.

See Figure 1A–D. A single I1/ (IPS1724a) is available from the sample. This tooth is completely preserved and only displays a very slight degree of wear, further showing a small contact facet at the apicalmost mesial side of the crown. It is a spatulate and relatively high tooth (labial height 7.0 mm), with the crown being progressively higher from distal to mesial. The moderately developed lingual cingulum is more marked at the mesial moiety of the crown than at the distal one, further forming a more developed ledge at the middle of the crown. There is a rounded thickening at the middle of the crown base below the cingulum, although it does not form a distinct pillar. The apical margin has distinct mammelons. In occlusal view, the crown base is asymmetrically elliptical, being convex both at the labial and lingual sides, but more markedly at the lingual one. In lateral view, the labial profile of the crown is convex, particularly at the apical half of the crown, whereas the lingual side is markedly concave apically from the cingulum. At the labial wall of the crown, a narrow crest runs from the cemento-enamel junction until almost reaching the apical margin.

Upper canine.

See Figures 1E,H, 2A–D, 6G–J, and 7A–E. The holotype specimen (IPS1724b) preserves the whole crown and the basal portion of the root, and merely displays a slight degree of wear with no dentine exposure. Given its relatively low crown height (7.1 mm on the buccal side) as compared to basal diameters, this tooth can be attributed to a female specimen. In occlusal view, the base of the crown and of the root is subtriangular and asymmetrical, particularly protruding at the lingual side, and not being markedly longer than wide (breadth/length index 82%). The buccal contour is slightly convex, whereas the lingual contour can be divided into a mesiolingual straight to slightly convex portion, and a slightly convex and somewhat shorter distolingual portion. In mesial and distal view, the buccal side of the crown displays a convex profile, whereas the lingual one is markedly concave above the cingulum. In buccal view, the crown displays an asymmetrical contour, with a slightly convex mesial profile and a rather concave (somewhat sigmoidal) distal one, so that the apex of the crown is somewhat tilted distally. At the lingual side of the crown, close to the cemento-enamel junction, there is a lingual cingulum that runs from the distalmost end to nearly the mesialmost end of the crown, forming a single stylar shelf that is quite wide and crescent-shaped. This shelf displays a marked lingual and occlusal thickening at the middle. There is a relatively sharp crest that runs mesially from the crown apex to the origin of the lingual cingulum, at the mesial end of the crown, near to the cemento-enamel junction. Just slightly lingually from this point, the cingulum is interrupted by a marked mesiolingual groove that runs parallel to the mesial crest by about its lower half. The distal crest that runs from the apex to the distal end of the crown (very close to the cemento-enamel junction) is slightly damaged at its middle portion, further displaying slight wear against the P/3, forming a sharp cutting edge with no dentine exposure.

The upper canine IPS1823 is further referred to the hypodigm of the newly described taxon, being attributed to a male individual on the basis of its larger size as well as some morphological differences (see below). The taxonomic affinities of this specimen have been widely discussed, being alternatively attributed to a male pliopithecid or a female hominoid. On the basis of morphological details (see Discussion), the former attribution is favored here, although given the above-mentioned discrepancies, we refrain from formally designating this specimen as a paratype of the new genus and species. This specimen is well preserved, except for some abrasion on the labial and distal portions of the crown, and includes the whole crown (buccal height 10.9 mm) and most of the root (except for its apicalmost portion; preserved root height 18.0 mm). This tooth displays a moderately advanced degree of wear: apical wear is minimal; and there is an elongated wear facet with no dentine exposure on the mesiobuccal aspect of the crown, a small wear facet with dentine exposure at the mesial end of the crown base, and a much larger wear facet with extensive dentine exposure on the distolingual portion of the crown. The root is much longer than the crown, and both are tilted distalwards on labial/lingual views. In occlusal view, the base of the crown displays a roughly elliptical profile, being somewhat longer than wide (breadth/length index 77%). The crown displays a slightly convex buccal contour, whereas the lingual one is divided into two distinct portions by a blunt crest that runs in lingual direction from the apex until close to the crown base. There are two sharp crests that run from the crown apex, in mesial and distal directions, thus further delimiting the mesiolingual and distolingual aspects of the crown, respectively. The mesiodistal portion of the crown is occupied by a broad and deep sulcus, which ends above an enamel thickening located on the mesial end of the base of the crown, which is mostly worn away (showing some dentine exposure). The distolingual aspect of the crown is markedly concave, at least on its apical moiety, since the basal and wider one displays a large wear facet against the P/3. The distal crest, although partially damaged, defines a sharp cutting edge that is further reinforced by the distolingual wear facet against the lower canine. No distinct cingula can be seen. As compared to the female C1/ from the holotype specimen, IPS1823 displays a higher crown (preserved buccal crown height 11.8 mm, i.e. 66% higher than in the female specimen), as well as longer and broader basal diameters (42 and 32%, respectively, since the male specimen is slightly more buccolingually compressed than the female one). Morphologically, IPS1823 further differs from IPS1724b by the much more marked and deeper mesiolingual sulcus, and the more elliptical (instead of subtriangular) occlusal outline; the latter is in turn related to the less buccolingually compressed morphology of the male crown, which unlike the female specimen further lacks a lingual cingulum.

Upper molars.

See Figures 4H,L, 5A–C, 11A–J, and 13A. A single M1/ is preserved (IPS1724c), including a complete crown with partial roots; the two lingual ones are fused into a single root (which is present), whereas the two buccal ones are broken away. This tooth displays a moderate degree of wear, with only a very slight dentine exposure at the apices of protocone, paracone and metacone. There are four main cusps, approximately equally developed, although the protocone is somewhat more extensive, and the metacone slightly more prominent, than the remaining ones. All the cusps are somewhat buccolingually compressed and pyramidal in shape (rather than being rounded and blunt), further being connected by relatively sharp crests. The crown displays a subrectangular outline, being somewhat buccolingually wider than mesiodistally longer (breadth/length index 123%). The buccal and distal contours are slightly convex, the lingual one is biconvex, and the mesial one is rather straight, although somewhat inclined, with the crown being longer at the buccal than at the lingual moiety. The lingual cusps are less peripherally situated than the buccal ones, which moreover are somewhat more mesially placed than the corresponding lingual cusps; this is particularly marked regarding the paracone, which is almost placed at the mesiobuccal corner of the crown. A markedly oblique and well-defined preprotocrista runs in mesiobuccal direction from the protocone, until forming a slight thickening at about the middle of the mesial margin of the crown, which can be interpreted as a vestigial paraconule. From this badly defined paraconule, both the lingual cingulum and the mesial marginal ridge originate in opposite directions. A relatively wide but very short preparacrista also originates in a mesial direction from the paracone, being continuous with the mesial marginal ridge. A quite restricted and shallow mesial fovea, much wider than long, is present, being mostly restricted to the buccal moiety of the crown. This fovea is mesially closed by the mesial marginal ridge, whereas distally it is separated from the much wider and deeper trigon basin by an ill-developed hypoparacrista that runs lingually from the paracone, until almost reaching the paraconule. As compared to the crown, the subtriangular trigon basin is not very wide, due to the quite medial position of the protocone. It is however somewhat larger than the distal fovea, which is slightly wider but appreciably shorter in mesiodistal direction. The trigon basin and the distal fovea are clearly separated by a fine crista obliqua, which is formed by a straight postprotocrista that runs from the protocone in distobuccal direction and an equally straight but somewhat longer hypometacrista that runs in mesiolingual direction from the metacone. The paracone and the metacone are united by two relatively wide crests that run in mesiodistal direction: a relatively long postparacrista and a much shorter premetacrista. The protocone and the hypocone are not clearly united by distinct crests; only a very short prehypocrista runs in a mesial direction, until merging with the postprotocrista before reaching the apex of the protocone. A fine and sharp posthypocrista runs in a distobuccal direction from the apex of the hypocone, originating the very fine but distinct distal marginal ridge, whereas an ill-defined crest that runs down from the apex of this cusp in a mesiolingual direction merges with the lingual cingulum. A wide but very short postmetacrista runs from the metacone in a distal direction, until merging with the buccal end of the distal marginal ridge. Although fine, this ridge closes distally the relatively deep distal fovea, which displays only a restricted development of secondary enamel wrinkling. Two distinct secondary crests, which are parallel to one another, can be appreciated at the distal fovea: a shorter one on the middle, and a longer and sharper one, which is more lingually placed. The latter secondary crest originates at the junction of the postprotocrista and the hypometacrista, running in a distolingual direction towards the base of the hypocone. This secondary crest, together with the postprotocrista, the prehypocrista, and the mesiobuccal wall of the hypocone's base, define a conspicuous although restricted triangle, distinct from both the trigon basin and the distal fovea. A distinct lingual cingulum is present, originating from the paraconule at almost the midline of the crown, and running backwards around the protocone until merging with the mesiolingual aspect of the hypocone. This lingual cingulum is hence absent from the distalmost third of the lingual wall of the crown, further being slightly interrupted at the level of the protocone by a short secondary crest that runs in lingual direction from this cusp. The buccal cingulum, in its turn, is less well marked, being mostly restricted to the region comprised between the paracone and the metacone, although with slight remains at the mesiobuccal aspect of the paracone and especially at the distobuccal aspect of the metacone.

Two (right and left) M2/ are also available from the sample (IPS1724d and IPS1724e), which preserve the complete crowns with partial roots; in particular, the left specimen preserves the fused, single lingual root, but the two buccal ones are broken, whereas the right tooth only preserves the distobuccal root. The degree of wear is slight, with no dentine exposure at all, although more marked than at the third molars. The crown is again slightly wider than longer (breadth/length index 123–127%), and the general shape of the crown and occlusal morphology is very similar to those of the first molar, only with some minor differences. Thus, there are four main cusps; the protocone is the most extensive, whereas the metacone is somewhat more prominent than the remaining ones. All these cusps are quite buccolingually compressed (especially the buccal ones), further being connected to one another by relatively sharp crests. The crown displays a subrectangular outline even more clearly than the first molar does. The distal and buccal contours are only slightly convex, whereas the mesial one is somewhat biconvex (rather than straight as in the M1/), and the lingual wall is markedly biconvex (much more than in the M1/). The lingual cusps are again more medially and less mesially situated than the buccal ones, although less markedly than in the M1/, particularly regarding the metacone, which is more distally situated. The occlusal pattern of the mesial portion of the crown is essentially comparable to the first molar, except that the preprotocrista and preparacrista are somewhat wider, whereas the mesial marginal ridge and the hypoparacrista remain relatively sharp. The mesial fovea is quite distinct, but again small, being much wider than long, and restricted to the buccal moiety of the crown. From the paracone towards the midline of the crown, the mesial fovea is somewhat curved in distolingual direction. On the specimen from the left side, this fovea is completely separated from the much deeper trigon basin by the somewhat irregular hypoparacrista, like in the M1/, whereas on the specimen from the right side, the latter crest ends next to—but does not reach—the paraconule, so that there is some connection with the trigon basin. The trigon basin is slightly wider and particularly longer than in the M1/, further being much wider and especially longer than the distal fovea, as compared to the first molar. The trigon basin and the distal fovea are separated by a sharp, straight and relatively well-developed crista obliqua. The paracone and the metacone are united by a marked continuous mesiodistal crest formed by the postparacrista and the premetacrista, which are partially interrupted (only in the specimen from the right side) by a very slight groove. Like in the M1/, the prehypocrista is not very distinct, being interrupted from the postprotocrista by a groove (more conspicuous in the specimen from the right side), which separates the distolingual aspect of the base of the protocone from the mesiolingual aspect of the base of the hypocone. There is no conspicuous posthypocrista distinct from the distal marginal ridge. The lingual cingulum similarly originates from a crest that runs in mesiolingual direction from the apex of the hypocone. Like in the M1/, there is a short, wide and quite indistinct postmetacrista, from which both the buccal cingulum and the distal marginal ridge originate. The lingual and buccal cingula are again well-developed, even more so than in the M1/. The lingual cingulum is wider and more ledge-like than in the M1/, but similarly originates from the paraconule, close the crown midline, running until the mesiolingual aspect of the hypocone. Like in the M1/, the lingual cingulum is interrupted at the lingual aspect of the protocone by a secondary crest; in the specimen from the right side, some additional secondary crests also radiate from the base of the protocone, although without disrupting the cingulum. The buccal cingulum further resembles that of the first molar; it is somewhat more continuous, being wider between the paracone and metacone, and being interrupted at the level of these cusps by some secondary enamel folds and crests. Like in the M1/, there is only a limited development of secondary wrinkling (enamel folds and crenulations) on the distal fovea. Unlike in the M1/, the triangle delimited by the postprotocrista and prehypocrista is not separated from the rest of the distal fovea by any particularly well developed secondary crenulation.

Finally, two (right and left) M3/ are also available (IPS1724f and IPS1724g), which preserve the complete crowns with broken roots, and which show no appreciable wear at all. The crown is much wider than longer (breadth/length index 127–130%), with a rather oval (instead of subrectangular) occlusal profile. The crown in wider at the mesial than at the distal moiety, and longer at the lingual than at the buccal moiety. In occlusal view, the outline of the crown is convex on the lingual and distal sides, whereas the mesial and buccal sides are straighter or slightly biconvex. There is a conspicuous swelling of the base of the crown, just above the cemento-enamel junction, at the distolingual corner (at the level of the hypocone), but it does not reach the occlusal level. The occlusal pattern significantly differs from that of the preceding upper molars. Thus, although the four main cusps are present, the protocone is by farthe most prominent and well-developed one, followed by the paracone, and then by the hypocone and metacone, the latter being quite rudimentary (almost vestigial). All the cusps are pyramidal in shape, and quite buccolingually compressed, especially in the case of the paracone. The various cusps are connected to one another by marked and relatively sharp crests, and the development of wrinkling (secondary crests and enamel folds) is more extensive than in the preceding upper molars, especially at the distal fovea. The protocone is more medially and distally situated than in the M1/ and M2/. The metacone is situated on the distobuccal corner of the crown, although like in the M2/ it is slightly more medially situated than the paracone and slightly less distally situated than the hypocone. The preparacrista, which runs mesiobuccally from the apex of the protocone, is wider but less distinct than in the preceding upper molars, whereas the paraconule is somewhat thicker. The mesial marginal ridge is distinct and well-marked, thus anteriorly closing the mesial fovea, which is somewhat larger than in the M1/ and M2/, but still much smaller than both the trigon basin and the distal fovea, and almost entirely circumscribed to the buccal moiety of the crown, as in the preceding upper molars. Like in the first and second molars, a relatively marked hypoparacrista closes more or less completely the posterior portion of the mesial fovea, running in a distolingual direction from the anterior base of the paracone to the anterior base of the protocone. The separation between the mesial fovea and the trigon basin, however, is less marked than in the preceding molars, which is accentuated by the fact that the trigon basin in the M3/ is shallower. In this tooth, the two mesial cusps are also somewhat more distally situated than in the preceding molars, so that the hypoparacrista originates at the level of the preparacrista (rather than from the apex of the paracone), and reaches the preprotocrista closer to the base of the protocone (well behind the paraconule). Moreover, unlike in M1/ and M2/, there is a secondary crest, parallel to the hypoparacrista but less well-defined, which runs from the apex of the paracone towards the center of the trigon basin. The trigon basin is approximately equal in size and shape to those of the preceding molars, although it appears larger as compared to the distal fovea, which is more restricted than in the M2/ and especially the M1/. The trigon basin and the distal fovea are completely separated from one another by the well-developed and straight crista obliqua. The apices of the paracone and metacone are joined by a moderately developed and slightly curved mesiodistal crest, formed by the postparacrista and the somewhat shorter premetacrista, which are continuous with one another. No single prehypocrista can be discerned uniting the hypocone with the protocone; instead, two crests originate from the hypocone in mesial and mesiobuccal directions, being directed towards the postprotocrista near the base of the protocone. No distinct posthypocrista can be discerned connecting the hypocone with the distal marginal ridge. In fact, the distal marginal ridge is discontinuous and variably developed, being interrupted by some secondary enamel crenulations, so that it does not completely closes the distal fovea. Similarly, but unlike in the preceding upper molars, there is no postmetacrista, since the metacone is directly situated at the junction of the mesial margin and the buccal cingulum, on the distobuccal corner of the crown. As a result, the distal fovea is triangular in shape, instead of subquadrangular or polygonal as in the preceding upper molars. The cingula, in its turn, are more developed than in the preceding upper molars, particularly the lingual one, which is wide, continuous and forms a C-shaped cingular shelf. This shelf displays several secondary crenulations that radiate from the base of the protocone, particularly on its mesiolingual aspect, but unlike in the preceding molars, none of these crests disrupts the cingulum. The buccal cingulum is at least as developed as in the M2/, running interruptedly from the paracone to the metacone. Besides the crenulations of the lingual cingulum, there is also some development of secondary tubercles and enamel wrinkling, particularly at the shallow distal fovea and, to a much lesser extent, at the trigon basin, thus contrasting with the condition displayed to this regard by the preceding upper molars.

Lower dentition

Lower canines.

See Figures 3A–H and 8A–J. C/1. Two (right and left) complete C/1 (IPS1724h and IPS1724i) are preserved. These include the crowns, which display only a very slight degree of wear, as well as the single roots, which are not yet closed. Given the general shape of these canines, as well as its restricted dimensions (buccal height 6.3 mm in the right one, and 6.4 mm in the left one), they can be attributed to a female individual. In occlusal view, the base of the crown displays a suboval (nearly elliptical) and considerably compressed contour, with the maximum (mesiodistal) diameter being much longer than the perpendicular (buccolingual) one (breadth/length index 63–65%), and being only minimally wider at the mesial than at the distal portion. In mesial and distal views, the crown is markedly asymmetrical: the single main cusp is slightly tilted towards the lingual side of the crown, and the buccal wall of the crown is markedly shorter than the lingual one, since the cementoenamel junction reaches highest at this point. In buccal or lingual view, the apex of the single main cusp is placed at about the middle of the crown, with the mesial contour being markedly convex and the distal contour being rather straight to slightly convex. From the crown apex, a relatively fine crest runs in mesiobuccal direction, until bifurcating somewhat below half the height of the crown. From this point, a fine mesial ridge descends in a mesial direction until reaching the mesialmost aspect of the crown's base, whereas a similarly developed but somewhat more irregular buccal cingulid runs in a distal direction slightly above the cemento-enamel junction. On the contrary, only a very restricted remnant of the lingual cingulid can be appreciated near to the distolingual corner of the crown. In occlusal view, the crown can be subdivided into three distinct portions. The lingual and mesial walls of the crown are continuous and markedly convex in occlusal contour, being separated from the flat and steeply inclined buccal wall by the already-mentioned crest that descends in mesiobuccal direction from the main cusp's apex. The distal wall of the crown is occupied by a distinct groove that is surrounded by two relatively sharp crests that descend from near to the crown apex to about the most basal third of the crown. The lingual crest converges with the remnant of the lingual cingulid at the distolingual corner of the crown, where a small rudimentary cusp can be found.

Lower premolars.

See Figures 4A–D and 9A–L. Three P/3 are preserved: the right (IPS1724j) and left (IPS1724k) specimens from the holotype, which preserve almost the complete crowns with no roots; and an additional right complete crown with partial roots (IPS34548, paratype). The left holotype specimen is only very slightly worn, and lacks the distalmost portion of the crown, whereas the right holotype specimen displays a comparable degree of wear (only slightly more worn at the tip of the main cuspid), further lacking the distolingual base of the crown. The paratype is more complete than the two holotype specimens, preserving not only the crown but part of the roots: apparently, a larger mesial root was present, being only partially fused with the more compressed distal root. The paratype is however more heavily worn than the two lower third premolars from the holotype, so that occlusal morphology can be better appreciated in the latter. In occlusal view, these teeth display a subtriangular contour (breadth/length index 65–73%), being only slightly wider towards the distal portion of the crown. There is a single main cuspid (protoconid), which is medially placed slightly on the mesial moiety of the crown. Like in the C/1, the crown can be subdivided into three distinct portions: a mesial and buccal portion, which is markedly convex in both lingual and occlusal views; a steeply inclined lingual portion, which is flat to slightly convex; and a restricted distal portion, which is markedly concave. The mesial and buccal portions of the crown are separated by a sharp crest (preprotocristid), which runs in a mesiolingual direction, until bifurcating at a small swelling situated at about one third of crown height. From this point, both the mesial marginal ridge (that runs until the mesialmost aspect of the crown) and the lingual cingulid originate. The distal portion of the crown is surrounded by two fine but well-defined short crests, which originate close but slightly below the crown's apex, and steeply descend in distobuccal and distolingual directions. The distobuccal crest (postprotocristid) terminates at a small thickening, where the distal marginal ridge and the buccal cingulid converge. The distolingual crest (hypoprotocristid) is shorter, ending in a small but individualized tubercle, which is placed at about half the crown's height (midway between the apex of the protoconid and the distolingual corner of the crown) and can be interpreted as a rudimentary but distinct metaconid. A lingual cingulid is present through most of the lingual aspect of the crown, but it is not very conspicuous, being more discernible at the mesial moiety of the crown, and ending below the above-mentioned tubercle just before reaching the distal marginal ridge. The buccal cingulid, on the contrary, is restricted to a somewhat wider but much shorter stylid at the distobuccal aspect of the crown. At the distal portion of the crown, there is a sharp and distinct marginal ridge.

Regarding the P/4, there is a single specimen (IPS1724l), which nearly preserves the whole crown except for the mesial wall and the basalmost portion of the buccal wall; it also partially preserves the two distal roots, which are fused with one another. This tooth is only moderately worn, with appreciable wear only at the tip of the two main cuspids, the protoconid and the metaconid. These cuspids are approximately equally extensive to one another, although the protoconid was apparently more prominent than the metaconid. An additional cuspid, the hypoconid, can be discerned at the distobuccal corner of the crown, although being much less extensive and less protruding than the two mesial ones, whereas the entoconid is reduced to a rudimentary thickening situated at the distolingual end, where the distal marginal ridge merges with a fine but distinct pre-entocristid. The buccal cuspids are somewhat more mesially situated that the corresponding lingual ones. The crown displays a subrectangular outline, being mesiodistally longer than buccolingually wide, and somewhat asymmetrical. In occlusal view, the buccal and lingual contours are somewhat convex, whereas the mesial and distal profiles are straighter. Two short but broad crests, the preprotocristid and premetacristid, run in a mesial direction from the protoconid and metaconid (respectively), presumably until reaching the mesial marginal ridge, which is not preserved. The mesial fovea (trigonid basin) is longer than wide, relatively deep and only slightly higher situated than the talonid basin; the former displays two secondary cristids that converge from the preprotocristid towards the center of the basin. Two irregular and poorly defined transverse crests, the hypoprotocristid and hypometacristid, which run towards each other from the protoconid and metaconid (respectively), distally close this mesial fovea, separating it from the deeper and much wider talonid basin. The latter is further delimited by two long but relatively indistinct crests, the postprotocristid and postmetacristid, which run in a distal direction from the protoconid and metaconid (respectively) until merging the much shorter prehypocristid and the above-mentioned pre-entocristid. The talonid basin is rectangular in shape (somewhat wider than long) and quite extensive in size, being delimited posteriorly by a well-developed distal marginal ridge, so that there is no distal fovea. There is no lingual cingulid, whereas the buccal cingulid is limited to a small buccal stylid just mesiobuccal from the hypoconid.

Lower molars.

See Figures 4E–G, 5D–F, 10A–I, and 13B,C. There is a single (left) M/1 (IPS1724m), which preserves both the crown and the intact roots (a mesial and a distal one, both with a bilobed cross-section). The crown displays a moderate degree of wear, with slight dentine exposure restricted to the apices of the cusps. There are five main cusps, which are rather pyramidal in shape and quite buccolingually compressed (especially in the case of the protoconid and hypoconid), instead of being blunt and rounded. The protoconid, metaconid, and hypoconid are approximately equally extensive, although the two mesial ones are more protruding (i.e., the trigonid is higher than the talonid). The entoconid, which is minimally damaged at the apex, is slightly less extensive and less protruding than the hypoconid, whereas the hypoconulid is by far the smallest (most restricted and least protruding) cusp of all. The hypoconulid is placed at the midline of the crown and next to its distal margin. From the remaining cusps, the mesial ones are more medially situated than the distal ones, resulting in a narrower trigonid basin as compared to the talonid basin. As a whole, however, the lingual cusps are more peripherally situated than the buccal ones, particularly the entoconid, which is placed near to the distolingual corner of the crown. Moreover, the two main buccal cusps are more mesially situated than the corresponding lingual ones, the difference between protoconid and metaconid being greater than the difference between hypoconid and entoconid. In occlusal view, the crown displays a subrectangular outline. The corners of the crown are markedly convex, but the margins are otherwise rather straight (especially the buccal one), except the lingual margin, which is moderately biconvex. The crown is markedly mesiodistally longer than buccolingually wide (breadth/length index 81%). The crown is however slightly asymmetrical, being somewhat longer in the buccal than in the lingual moiety, due to the very mesial placement of the protoconid. On the contrary, the mesial and the distal lobes of the tooth are approximately equally wide at the base of the crown. A short but wide preprotocristid runs from the apex of the protoconid in a mesial direction, until reaching the mesial margin of the crown. A longer and more irregular premetacristid, in its turn, curves in mesiobuccal direction from the metacone, until merging with the mesial marginal ridge, which anteriorly delimits the mesial fovea (or trigonid basin). This basin is relatively narrow as compared to total crown width, and much shorter and narrower than the talonid basin, although still quite extensive, displaying a subquadrangular outline (not being appreciably longer than wider). The trigonid basin is separated from the much deeper talonid basin by two fine crests, the hypoprotocristid and the somewhat longer hypometracristid, which descend in buccolingual direction from the distal aspect of the apex of the protoconid and from the medial aspect of the metaconid apex, respectively. A short and quite indistinct postprotocristid further originates from the distal aspect of the base of the protoconid, whereas a similarly short prehypocristid runs in a mesial direction from the hypoconid. These crests almost make contact halfway between the protoconid and hypoconid, being only minimally separated by a very fine transverse groove. Together, the postprotocristid and the prehypocristid define a moderately inclined cristid obliqua, which also constitutes the lateral margin of the pliopithecine triangle. Another crest, quite wide and distinct, further runs in mesiolingual direction from the mesial aspect of the hypoconid, constituting the distal arm of the pliopithecine triangle. The latter, however, is not completely formed, since the mesial arm is lacking, being substituted by a groove situated where the mesial arm of the triangle should be. A simple pattern of fine grooves is present at the deepest parts of the talonid and trigonid basins, separating the bases of the several cusps from one another. The lingual margin of the talonid basin is defined by a fine postmetacristid, which runs in distal direction from the base of the metaconid, together with a fine pre-entocristid, which runs from the apex of the entoconid in mesiolingual direction, until nearly contacting with the end of the postmetacristid, from which it is separated by a fine groove. Finally, the talonid basin is closed distally by the medially positioned hypoconulid, as well as by the very short and ill-defined crests that connect it with the surrounding cusps: the posthypocristid and prehypoconulid cristid, in the case of the hypoconid; and the hypoentocristid and postcristid, in the case of the entoconid. The hypoentocristid, however, can be barely discerned, further being separated from the end of the short but more distinct postcristid by a narrow groove that runs onto the distal fovea. The latter is quite restricted, due to the medial position of the hypoconulid and the distal position of the hypoconid, thus being entirely circumscribed to the lingual moiety of the crown. There is no lingual cingulid except by a small stylid at the level of the postmetacristid and pre-entocristid, and some additional subtle remnants around the metaconid. The buccal cingulid is somewhat more developed, although still restricted and discontinuous. The first portion runs from the mesial aspect of the protoconid to the mesiobuccal aspect of the metaconid, being particularly developed in between these two cups. The second portion of the buccal cingulid includes an additional stylid, which is situated between the hypoconid and hypoconulid at the distobuccal corner of the crown. The development of secondary enamel crenulations is very restricted.

There is also a single (right) M/2 (IPS1724n), which preserves the whole crown without roots. The degree of wear is very slight, being limited to the tips of the four more protruding cusps, but with no dentine exposure. In general, the occlusal morphology of this tooth is very similar to that of the first lower molar, and the crown as a whole is similarly narrow in relative terms (breadth/length index 80%), although absolutely larger. Like in the M/1, the mesial cusps (particularly the protoconid) are more prominent than the remaining ones, while the hypoconid is the most voluminous one. The hypoconulid is again the least developed cusp, although it is comparatively larger than in the M/1. The disposition of the cusps is the same, with the hypoconulid positioned at the middle of the distal margin of the crown, and the two other buccal cusps (particularly the protoconid) being more medially positioned than the corresponding lingual cusps. The crown displays a subrectangular to elliptical outline, with more markedly convex profiles than in the M/1. The crown of the M/2 is similarly longer in the buccal than in the lingual moiety of the crown, but differs from the preceding molar by being somewhat wider at the mesial than at the distal lobe. The mesial fovea (trigonid basin) is well delimited (especially on its mesial aspect, by a marked mesial marginal ridge) and quite developed, although being narrower relative to the width of the crown compared to the M/1. Due to the somewhat greater extension of the cusps as compared to the preceding molar, the protoconid almost reaches the mesial margin of the crown, so that there is no distinct preprotocristid. On the contrary, a relatively wide premetacristid runs in a mesiobuccal direction, soon becoming continuous with the mesial marginal ridge. The hypometacristid, which separates the mesial fovea from the talonid basin, is relatively wide but somewhat indistinct, although less so than the shorter hypoprotocristid, which runs in the opposite direction from the apex of the protoconid. These two crests do not join at the middle, as in the M/1, but are separated by a fine groove that runs in a mesiodistal direction, thus minimally allowing communication between the mesial fovea and the deeper talonid basin. Like in the preceding molar, the pliopithecine triangle is not completely formed, although it can be more clearly discerned than in the M/1. Two short but wide mesiodistal crests (postprotoconid and prehypocristid), which constitute the slightly inclined cristid oblique, are separated from one another by a fine transverse groove. They also constitute the external side of the triangle, which is not completely constituted because the mesial arm is substituted by a tubercle; moreover, the distal arm of the triangle is lingually (instead of mesiolingually) directed, thus being separated from the above-mentioned secondary cuspulid by a wide groove. The postmetacristid, which runs in distolingual direction from the metaconid, is more well-defined than in the M/1. This crest is directly separated from the base of the entoconid by a buccolingual groove, since no distinct pre-entocristid is present. The talonid basin is subrectangular, being slightly longer than broad, although its deepest part is more clearly V-shaped than in the M/1, due to the greater extension of the hypoconid and of the distal arm of the pliopithecine triangle. On the lingual side, the talonid basin is delimited by the postmetacristid and the entoconid, whereas on the buccal side, it is delimited by the pliopithecine triangle and the hypoconid in a strict sense. Distally, like in the M/1, this basin is closed by the hypoconulid. Although there are no clearly distinct posthypocristid or prehypoconulid cristid, the bases of the hypoconid and the hypoconulid contact with one another, thus completely closing the talonid basin on its distal portion. On the contrary, the contact between the short and quite indistinct postcristid and the hypoentocristid (which can be barely distinguished from the base of the entoconid) is interrupted by a fine groove, like in the M/1, thus partially allowing communication between the deep talonid basin and the much shallower distal fovea. The latter is similarly situated close to the distolingual corner of the crown, although it is somewhat larger than in the M/1, due to the less distally positioned entoconid, resulting in a diffuse but appreciable postentocristid that runs in distal direction. There is no discernible lingual cingulid, whereas on the contrary, the buccal cingulid is more well-developed than in the M/1. Nevertheless, the latter cingulid is still discontinuous, being divided by the hypocone into two distinct portions: a long and continuous portion, running from the mesial aspect of the protoconid until the mesiobuccal aspect of the hypoconulid, and becoming ledge-like between the protoconid and the hypoconid; and a shorter but similarly well-developed stylid, which is situated at the distobuccal corner of the crown, between the hypocone and the hypoconulid. The development of secondary enamel wrinkling, like in the M/1, is very restricted, mainly consisting of several crenulations situated on the lingual aspect of the base of the hypoconid.

Finally, there is a single M/3 (IPS1724o), which completely preserves the crown but lacks the roots, except for a very small portion of the mesial one. This tooth shows no wear at all. Its occlusal morphology differs from both the M/1 and M/2 in several respects, including cusp development, position of the hypoconulid, distinctiveness of the pliopithecine triangle, and development of secondary enamel folds and wrinkles, among others. The two mesial cusps (particularly the metaconid) are more protruding than the remaining ones. The metaconid is also the most extensive cusp, followed by the protoconid and hypoconid, and then by the entoconid (more reduced than in the preceding lower molars) and the hypoconulid, with the latter being restricted to a rounded small tubercle. Besides being comparatively less developed than in the preceding molars, the hypoconulid is more buccally positioned, being situated on the buccal moiety of the crown, although not completely aligned with the remaining buccal cusps. Like on the preceding molars, the lingual cusps (especially the entoconid) are more peripheral than the buccal ones. The crown displays a subrectangular to suboval outline, being much longer than broad (breadth/length index 70%), and with the maximum crown breadth being attained at the mesial lobe. The occlusal profile of the crown is somewhat asymmetrical, but unlike the preceding molars, the lingual moiety of the crown is longer than the buccal one, due to a much more extensive development of the distal fovea. In occlusal view, the crown is moderately convex at the mesial and buccal sides, whereas it is highly biconvex at the lingual side, and markedly convex at the distal one. The mesial fovea (trigonid basin) is very well delimited and even more extensive than in the preceding lower molars, being somewhat wider than long. There is a very short, but wide and distinct preprotocristid, while a finer but longer premetacristid curves in mesiobuccal direction until merging with the mesial marginal ridge, which is quite distinct and mesially closes the mesial fovea. Distally, this fovea is closed by a fine but relatively sharp transverse crest, formed by the short hypometacristid and the longer hypoprotocristid. Besides the slightly inclined cristid obliqua (formed by the short and fine postprotocristid, and the fine but longer prehypocristid, which do not completely merge with one another), the pliopithecine triangle is well-delimited by two (mesial and distal) distinct and sharp arms. Accordingly, the pliopithecine triangle is much better developed than on the preceding molars, further being completely separated from the rest of the talonid basin because the two (mesial and distal) arms contact with one another on their medial ends. Lingually, the talonid basin is surrounded by the more or less distinct postmetacristid that runs in distal direction from the metaconid, as well as by the entoconid, since there is no distinct pre-entocristid; instead, as in the M/2, the base of the entoconid is separated from the end of the postmetacristid by a relatively fine buccolingual groove. The talonid basin of the M/3 displays a polygonal shape, and unlike in the M/1 and M/2, it is not closed by the hypoconulid, which is much more buccally positioned than in the preceding molars. The hypoconid and the hypoconulid are connected with one another by two short and broad crests, the posthypocristid and the more indistinct prehypoconulid cristid, which further separate the talonid basin from the buccal wall of the crown. The hypoconulid is situated close to the distobuccal aspect of the crown, where both the lingual cingulid and the distal marginal ridge converge, since there is no distinct posthypoconulid cristid. However, a distinct and well-defined postcristid runs in mesiolingual direction from the hypoconulid towards the entoconid, thus partially separating the talonid basin from the distal fovea. This crest, however, ends before reaching the distinct but much shorter hypoentocristid that runs in distobuccal direction from the entoconid. Accordingly, unlike in the preceding lower molars, and in spite of the more developed postcristid, the distal fovea of the M/3 largely communicates with the talonid basin. Moreover, secondary enamel wrinkling is much more marked on both the distal fovea and the talonid basin, including the presence of several secondary crests that are very well-developed. Two of these secondary crests originate from the end of the hypoentocristid. The distal one curves backwards until merging with the short postentocristid, which is continuous with the distal marginal ridge, thereby isolating a small and rounded portion of the distal fovea. The mesial one, on the contrary, runs in buccal direction, progressively curving backwards until contacting with the end of the short posthypocristid, thus completely separating the largest and deepest portion of the talonid basin from a smaller and shallower portion, which is not completely isolated from the distal fovea. In other words, unlike in the preceding molars, the talonid basin is subdivided into three distinct portions: the well-constituted pliopithecine triangle; the largest and deepest portion, distally delimited by the above-mentioned transverse secondary crest; and the distalmost portion, which is not completely separated from the distal fovea. Two additional secondary crests are also present on the distal fovea: a narrower one, perpendicular to the postcristid; and a wider, tubercle-like one, situated next to the distolingual corner of the crown. The distal marginal ridge is distinct and markedly curved, posteriorly closing the distal fovea, which is larger than in the preceding lower molars, not being entirely restricted to the lingual moiety of the crown. Unlike in the M/1 and M/2, there is no lingual cingulid or any remnant of it. On the contrary, a nearly continuous buccal cingulid is present from the mesial aspect of the protoconid to the distal aspect of the hypoconulid, being of comparable width to that in the M/2, but being only partially interrupted at the level of the hypoconid.

METRICAL AND MORPHOLOGIC COMPARISONS

Dental size and proportions

The dental size (length and breadth) of Barberapithecus huerzeleri sp. nov., as compared to other pliopithecid taxa, is depicted in Figures 14 and 15, respectively, regarding the lower and upper canines and cheek teeth. The lengthand proportions of the lower molars of this taxon, compared to other small-sized pliopithecids, are reported in Figure 16, whereas the relationship between the lower premolar and lower molar series is depicted in Figure 17. Metrically, the upper molars of Barberapithecus most closely resemble Pliopithecus canmatensis and, to a greater extent, Dionysopithecus, although it must be stressed that the condition is unknown for Plesiopliopithecus (Fig. 15). Regarding the lower dentition (Fig. 16), the female lower canines of Barberapithecus resemble in size and proportions those of Pliopithecus canmatensis (condition unknown for Plesiopliopithecus and Dionysopithecus), whereas the lower cheek teeth also most closely resemble P. canmatensis, Dionysopithecus and, to a greater extent, Plesiopliopithecus (Fig. 14). Barberapithecus, like P. canmatensis, further differs from Dionysopithecus by displaying an M/3 absolutely longer and relatively narrower than the M/2 (Fig. 16); the former condition is further displayed by other Pliopithecus species (with the exception of M/3 of P. piveteaui, which is relatively narrower but not absolutely shorter than the M/2), and cannot be ascertained in Plesiopliopithecus. However, all the lower molars of Barberapithecus are relatively narrower than those of Pliopithecus spp. and Dionysopithecus, although displaying an apparently intermediate condition between other crouzeliines and both pliopithecines and dionysopithecines (Fig. 16). Barberapithecus also resembles Dionysopithecus and P. canmatensis by displaying a short premolar series relative to the molar series (unlike in P. antiquus and Epipliopithecus; condition unknown in Plesiopliopithecus; Fig. 17A), although the relative lengthbetween the P/4 and the M/2 further permits to distinguish Barberapithecus, P. canmatensis, and Plesiopliopithecus from Dionysopithecus, the latter displaying, like P. antiquus, a shorter P/4 as compared to M/2 length (Fig. 16B).

Figure 14. Bivariate plots of buccolingual breadth (BL) versus mesiodistal length (MD) in lower canines and cheek teeth of Barberapithecus huerzeleri sp. nov. from Castell de Barberà. Measurements taken from originals in the case of B. huerzeleri, and from casts and published sources regarding the other taxa: Hürzeler (1954); Zapfe (1960,1961); Ginsburg (1975); Ginsburg and Mein (1980); Pan et al. (1989); Harrison et al. (1991); Welcomme et al. (1991); Harrison and Gu (1999); Moyà-Solà et al. (2001); Alba et al. (2010).

Figure 15. Bivariate plots of buccolingual breadth (BL) versus mesiodistal length (MD) in upper cheek teeth of Barberapithecus huerzeleri sp. nov. from Castell de Barberà. See Figure 12 for further details on the measurement sources. Measurements taken from originals in the case of B. huerzeleri, and from casts and published sources regarding the other taxa: Hürzeler (1954); Zapfe (1960); Ginsburg (1975); Pan et al. (1989); Harrison et al. (1991); Harrison and Gu (1999); Alba et al. (2010).

Figure 16. Lower cheek-teeth length and proportions in Barberapithecus huerzeleri gen. et sp. nov. from Castell de Barberà, as compared to other, similarly sized pliopithecids: (A) Mesiodistal length (in mm); (B) Breadth/length index (%). Measurements taken from originals in the case of B. huerzeleri, and from casts and published sources regarding the other taxa: Hürzeler (1954); Zapfe (1960); Ginsburg (1975); Pan et al. (1989); Harrison et al. (1991); Harrison and Gu (1999); Moyà-Solà et al. (2001); Kordos and Begun (2001); Alba et al. (2010).

Figure 17. Lower dentition proportions in Barberapithecus huerzeleri gen. et sp. nov. from Castell de Barberà, as compared to other pliopithecids: (A) lower premolar series length versus lower molar series length; (B) mesiodistal length (MD) of the P/4 versus that of the M/2. The length of each series was computed as the sum of average mesiodistal length for the several teeth (in mm). Measurements taken from the same sources as in Figure 14.

MORPHOLOGIC COMPARISONS

Upper dentition

The I1/ of Barberapithecus is very similar to that of Pliopithecus, differing from both Dionysopithecus and Anapithecus by the higher, mesiodistally shorter and relatively labiolingually narrower crown, and from Anapithecus also by the less prominent lingual cingulum. As compared to other anapitheciins, the pliopithecine-like incisor morphology of Barberapithecus can be therefore considered plesiomorphic. On the contrary, the C1/ of Barberapithecus is autapomorphic, with its main cusp being in both sexes more buccolingually compressed than in pliopithecines (except for Epipliopithecus) and other crouzeliines. Thus, the female C1/ of Barberapithecus is very distinctive, because its main cusp is markedly buccolingually compressed and mesiodistally elongated, further displaying well-developed mesiodistally oriented crests. The crown base is buccolingually much wider than the main cusp, with an extensive lingual shelf that is lacking in Pliopithecus platyodon and P. canmatensis, where the crown base is only subequal in size to the base of the main cusp. The female C1/ of Anapithecus further displays a more rounded base and less compressed main cusp, as well as a reduced mesial sulcus and a better-developed lingual cingulum, as compared to Barberapithecus. The male C1/ of the latter is also very distinctive, with the crown being more buccolingually compressed than in Pliopithecus platyodon but not than in Epipliopithecus vindobonensis (see next section for additional discussion regarding the taxonomic attribution of the male canine described in this article).

Regarding the upper molars, Barberapithecus differs from those of dionysopithecines in many respects. Thus, the upper molars of the new genus differ from those of Dionysopithecus by the more buccolingually compressed cusps, the less oval contour that is not slightly shorter lingually than buccally, the more peripheral protocone (instead of situated close to the crown midline), the paraconule situated farther away from the protocone, the shorter hypoparacrista and more restricted (particularly narrower) mesial fovea (due to more median position of the paraconule in Barberapithecus), the hypocone not situated more lingually than the protocone, the more developed lingual cingulum that nevertheless does not reach the apex of the hypocone, and the presence of a distinct (although short) posthypocrista. As compared topliopithecines, the M1/ crown of Barberapithecus is relatively narrower than usual in Pliopithecus and Epipliopithecus, and displays less inflated and more buccolingually compressed cusps, which are longer than wide (whereas in pliopithecines, on the contrary, the cusps are wider than long, in particular the labial ones), as well as more well-developed and sharper crests. The labial cingulum in Barberapithecus displays a relatively high position, thus contrasting with the low-situated cingulum of pliopithecines, particularly P. canmatensis. Barberapithecus further differs from P. canmatensis, P. platyodon, and Epipliopithecus by the more peripheralized protocone and hypocone, as well as by the long preprotocrista connecting the protocone to the paraconule, and the short and mesially curved hypoparacrista connecting the paracone to the paraconule, thus defining a narrow mesial fovea, with the paraconule nonperipherally situated, closer to the paracone. Moreover, both the protocone and hypocone are more distally situated than in the above-mentioned taxa, with the latter cusp being close to the distolingual corner of the crown in Barberapithecus, instead of being approximately aligned with the apex of metacone. The M1/ and M2/ of Barberapithecus further differ from those of the above-mentioned taxa by the reduced lingual cingulum (which in Epipliopithecus does not reach the apex of the hypocone), and from Epipliopithecus by the lack of buccal cingulum. The M1/ and M2/ of Barberapithecus most closely resemble the occlusal morphology displayed by those of anapithecins, particularly Anapithecus (unknown in crouzeliins), although with some differences. Thus, Barberapithecus resembles Anapithecus by the long preprotocrista, the somewhat distally situated protocone, the short hypoparacrista, and the restricted distal fovea. However, Barberapithecus differs from the type genus of the Anapithecini by the less buccolingually compressed cusps, the less well-developed and less sharp crests, the more distally situated hypocone (farther away from the protocone), the better-developed lingual cingulum in the M1/ and M2/ (reduced in Anapithecus to the mesial and lingual aspects of the protocone), and the shorter and wider M3/.

Lower dentition

No lower incisors of Barberapithecus are available. Regarding the female C/1 of the newly erected taxon, like the upper canine, displays autapomorphic morphology, being extremely buccolingually compressed. Moreover, the alignment of the mesiodistal axis of the crown (as defined by the contact facet against the P/3 and the crown apex) suggests that this tooth was inserted into the mandible less obliquely in respect to the postcanine toothrow than in other pliopithecids.

With regard to the lower premolars, Barberapithecus differs from most pliopithecids by the presence of a distinct (although small) metaconid at the end of the hypoprotocristid, which is a feature only present in Anapithecus, Egarapithecus and Laccopithecus (although variable in the latter; see Moyà-Solà et al.,2001), and also to a lesser extent in Dionysopithecus (although in the latter taxon the metaconid is not a truly individualized cusp, since the whole hypoprotocristid is inflated along its total length; Harrison and Gu,1999). The maximum development of the P/3 metaconid is attained in Egarapithecus (Moyà-Solà et al.,2001), whereas on the contrary this feature is absent in both Pliopithecus and Plesiopliopithecus, and Epipliopithecus only shows at most a rudimentary swelling that cannot be interpreted as an individualized cusp. The P/3 of Barberapithecus further displays a subtriangular occlusal outline (somewhat wider distally), most closely resembling the female P/3 of Anapithecus (although being less inflated), and markedly contrasting with the contour of the female P/3 of Pliopithecus, which are more buccolingually compressed and tend to be wider mesially. The P/4 of Barberapithecus is unique amongst pliopithecids in several features, such as displaying a trigonid and talonid of similar length, whereas in other pliopithecids for which this tooth is known the trigonid is to some extent shorter than the talonid; the presence of a less elongated P/4 talonid in Barberapithecus compared to other anatopitheciins may be considered a plesiomorphic feature. Moreover, the talonid basin is shorter than wide, with small and low hypoconid and entoconid; the presence of the latter cusps is unusual but not unknown amongst pliopithecids (e.g., Pliopithecus canmatensis; Alba et al.,2010). The P/4 of Barberapithecus is further characterized by the presence of a well-developed premetacristid originating from the metacone apex and lingually (probably also mesially) closing the trigonid basin (mesial fovea). This is a derived feature present in most crouzeliines (Plesiopliopithecus and Anapithecus), but absent in Egarapithecus and pliopithecines; in Laccopithecus, there is a premetacristid, although it originates from the base instead of from the apex of the metaconid (as in Dionysopithecus).

Finally, regarding the lower molars, the M/1 and M/2 of Barberapithecus differ from those of pliopithecines by the more peripheralized and buccolingually compressed cusps, the larger mesial and distal foveae, the more quadrangular and restricted central fovea, and the mesially situated entoconid and the distally situated hypoconid with a short (with these two cusps approximately aligned with one another); moreover, in the M/2 of Barberapithecus the distance between the protoconid and metaconid is only slightly shorter than between the hypoconid and entoconid. By these features enumerated above, Barberapithecus most closely resembles other crouzeliines, especially Anapithecus, although it differs from the latter by the larger distal fovea and the more mesially situated entoconid in the M/2 (instead of being more posteriorly situated, as in Anapithecus and pliopithecines). The M/3 of Barberapithecus resembles the M/2 of the same taxon by displaying peripheralized cusps, a wide and high trigonid, a large mesial fovea and a quadrangular central fovea that is not longer than wide, a distally situated metaconid relative to the proconid, a small and low entoconid that is slightly mesially situated, and a small hypoconulid. In several of these regards, Barberapithecus clearly differs from Pliopithecus species, including the more peripheralized cusps, the wider, deeper and higher trigonid, the larger mesial fovea, the nonelongated central fovea, the smaller and more mesially situated entoconid, the smaller hypoconulid, and the fact that the metaconid is more distally situated than the protoconid (thus resembling P. platyodon and P. canmatensis, and differing from P. piveteaui and P. antiquus), as well as the comparatively somewhat reduced distal lobe in the former. In most of these and other features, Barberapithecus rather resembles the condition displayed by Anapithecus, including the buccolingually compressed cusps and crests, the peripheralized cusps, the distally situated metaconid as compared to the protoconid, the wide and high trigonid, the largeanterior fovea, and the small and low entoconid. Nevertheless, the M/3 of Barberapithecus differs from Anapithecus by the possession of less elongated (less relatively narrow) crown and central fovea, the less peripherally situated hypoconulid, and the more developed buccal cingulid.

DISCUSSION

Number of preserved individuals

Crusafont-Pairó and Golpe-Posse (1981) attributed the sample from CB to a single individual, since all the remains were found “together” (no further taphonomical or spatial information is provided) and because they all display a “slight and similar” degree of wear. According to Moyà Solà et al. (1990) and Andrews et al. (1996), on the contrary, at least two different individuals would be represented in the sample: a subadult female individual (including most of the material), and a second one, being represented by an isolated P/3. Crusafont-Pairó and Golpe-Posse (1981), however, listed and described all the lower teeth from CB known at that time, and only mentioned two (right and left) lower third premolars, which they attributed to the same individual. Most of these teeth were found by the late Narciso Sánchez (employee of the former Institut de Paleontologia de Sabadell during the 1970's) in semi-anatomical connection within a single block of green clay, which came from the same classical level of CB that has produced most of the faunal remains from this site (Narciso Sánchez, pers. com. to SMS). Initially, only 14 teeth from CB were reported (Anonymous,1975b), so that the fifteenth tooth included in the holotype must be the lower molar found by M.Crusafont and J.M. Golpe somewhat later than the original discovery at the same locality (Anonymous,1975a). The additional P/3 from a second individual, here described as a paratype, was found several years later by one of the authors of this article (SMS) among the fossil collections from CB, then housed at the former Institut de Paleontologia M. Crusafont. Accordingly, there are solid grounds for arguing that all the teeth here included in the holotype of Barberapithecus huerzeleri gen. et sp. nov. belong to a single individual.

We have further included in the description of the new genus and species the upper male canine IPS1823 from CB (IPS26 in the old nomenclature employed by Golpe Posse,1982,1993), whose affinities have been widely discussed. Several authors (e.g., Harrison,1991; Golpe Posse,1993) considered in the past that this specimen was a female upper canine of Hispanopithecus laietanus (or Dryopithecus laietanus), but more recently Begun (2002) and Alba et al. (2010) concluded that this specimen corresponds instead to a larger individual of the same pliopithecid taxon. We agree with the latter view, given the morphological differences between this tooth and the upper female canines of Hispanopithecus laietanus (see Fig. 2). For example, in buccal view the root of IPS1823 is distinctly mesiodistally curved and the crown is more distinctly triangular than in the female upper canines of Hispanopithecus. Given that the size discrepancy between the former specimen and the upper canine from the holotype can be easily accounted for by sexual dimorphism, we tentatively interpret IPS1823 as a male canine of the newly erected pliopithecid taxon. However, given the fact that a large-bodied hominoid is also present at Castell de Barberà on the basis of postcranial material (Alba et al.,2011b; Almécija et al.,2011), the possibility cannot be completely discarded that IPS1823 corresponds to a male canine of some other hominoid taxon. This is the reason why we have refrained from formally designating this specimen as a paratype of Barberapithecus huerzeleri, even though we refer it to the hypodigm of the latter.

Begun (2002) further argued that the isolated right dP/4 from Can Feliu (=Can Sant Feliu; IPS34565), mistaken for a M/3 by Golpe-Posse and Crusafont (1981), and attributed to aff. Crouzeliinae indet. by Moyà Solà et al. (1990) and to Crouzeliinae indet. by Andrews et al. (1996), might belong to another larger individual of the same taxon from CB. Given the presence of other pliopithecid taxa in the Vallès-Penedès Basin during the Miocene—including Pliopithecus canmatensis in the Abocador de Can Mata series (Alba et al.,2010) and Pliopithecus sp. in Sant Quirze (Harrison et al.,2002)— the comparatively larger dental size of the taxon represented at Sant Feliu (at least, as compared to P. canmatensis and the CB taxon), and the lack of homologous material, for the moment being we refrain from providing a formal taxonomic attribution for this specimen.

Systematic position and phylogenetic affinities

The most outstanding morphologic features of Barberapithecus huerzeleri gen. et sp. nov. are the buccolingually compressed cusps, the well-developed and sharp crests, and the large and well-defined occlusal foveae. This is the reason why Crusafont-Pairó and Golpe-Posse (1981,1982) initially noted similarities between the pliopithecid material from CB and that currently attributed to the crouzeliine Plesiopliopithecus lockeri, and somewhat later both Ginsburg (1986) and Moyà-Solà et al. (1990) recognized possible crouzeliine affinities for the former. Later, however, Andrews et al. (1996) concluded that the material from CB most likely corresponded to Pliopithecus antiquus. These authors recognized that the material from CB was smaller than the male holotype of P. antiquus from Sansan, but attributed this fact to sexual dimorphism, noting further size similarities with the smaller mandibular fragment from La Grive. Andrews et al. (1996) also pointed out to some morphological differences, with the CB taxon displaying narrower lower molars, but they considered such differences merely attributable to interspecific variation. Somewhat later, Begun (2002) argued that the taxon from CB is smaller and morphologically distinctive from P. antiquus, as indicated by the presence of laterally compressed cusps, well-developed crests, and large occlusal basins. This author asserted that the CB taxon is more similar to Pliopithecus piveteaui, but further displaying some differences that would justify attribution to a new pliopithecine taxon. Most recently, Alba et al. (2010) similarly asserted that the taxon from CB cannot be attributed to either P. antiquus or the newly described species Pliopithecus canmatensis, merely attributing the former to Pliopithecidae nov. These discrepancies on the systematic affinities of the CB taxon stem from the peculiar association of crouzeliine, pliopithecine and autapomorphic features, which make it difficult to decipher its closest phylogenetic affinities and, hence, it is most suitable systematic placement.

Besides the above-mentioned dental features of Barberapithecus, which distinguish the overall occlusal pattern of crouzeliines from that of other pliopithecids, the newly erected genus from CB displays a set of occlusal details that further support a close phylogenetic relationship with crouzeliines, and particularly Anapithecus: upper molars with peripheralized protocone and hypocone, long preprotocrista, short and mesially curved hypoparacrista, short distal fovea and somewhat distally situated protocone; P/3 with a distinct metaconid at the end of the hypoprotocristid; P/4 with a well-developed premetacristid that lingually closes the trigonid; deep trigonid foveae in the P/3 and lower molars; and lower molars with more peripheralized cusps, and a relatively large and deep mesial fovea, with the distance between the apices of the protoconid and metaconid in the M/2 and M/3 being only slightly smaller than between those of the hypoconid and entoconid, with the latter cusp being small and not very protruding. On the contrary, Barberapithecus is morphologically more similar to pliopithecines regarding other occlusal details (see Figs. 12 and 13): the I1/ is high and narrow; the metaconid and the protoconid of the M/1 are more closely situated than the entoconid and the hypoconid, so that the trigonid basin is narrower than the central fovea; there is a large and centrally situated hypoconulid in both the M/1 and M/2; and there are well-developed cingula in both the upper and lower molars (although somewhat reduced as compared to pliopithecines). Finally, Barberapithecus displays several unique features amongst pliopithecids, including: female C1/ with a markedly buccolingually compressed and proximodistally elongated main cusp, with an extensive lingual shelf; female C/1 extremely buccolingually compressed and mesiodistally oriented; P/4 with a large trigonid, subequal in size to the talonid; lower molars with quadrangular central foveae, mesially situated entoconid, and distally situated hypoconid that is situated more or less in front of the entoconid in the M/2 and M/3; short posthypocristid; very large distal fovea in the M/1 and especially the M/2; reduced distal lobe of the M/3.

Although some disagreements persist among authors regarding the composition of the Crouzeliinae, with Egarapithecus being considered a crouzeliine by several researchers (Moyà-Solà et al.,2001; Harrison et al.,2002; Alba et al.,2010) but a pliopithecine by some (Begun,2002), there seems to be a consensus (e.g., Begun,2002; Alba et al.,2010) that the Crouzeliinae constitute a monophyletic group, composed at least by the genera Plesiopliopithecus, Anapithecus, and Laccopithecus, and defined by several occlusal synapomorphies. Given the mosaic of crouzeliine and pliopithecine dental features displayed by Barberapithecus, it must be stressed that similarities with the former are considered shared-derived derived features defining the Crouzeliinae (Harrison and Gu,1999; Moyà-Solà et al.,2001; Begun,2002), whereas the pliopithecine features of the newly erected taxon are most likely plesiomorphic features. Interestingly, the breadth/length index of the lower molars of Barberapithecus is intermediate between that of dionysopithecines plus pliopithecines, on the one hand, and that of crouzeliines (Plesiopliopithecus), on the other, therefore suggesting that Barberapithecus is also somewhat derived towards the crouzeliine condition regarding dental proportions. Taken overall, therefore, the derived features shared between Barberapithecus and other crouzeliines are most likely interpreted as indicating a close phylogenetic link between these taxa. Leaving Barberapithecus aside, two distinct groups can be distinguished amongst crouzeliines based on both occlusal morphology and stratigraphic range. The genus Plesiopliopithecus (=Crouzelia)—here included into the monotypic tribe Crouzeliini—includes several poorly known species from the European Middle Miocene (Zapfe,1961; Bergounioux and Crouzel,1965; Ginsburg and Mein,1980; Begun,2002): Pl. lockeri from Trimmelkam (MN6); Pl. auscitanensis from Sansan (MN6); and Pl. rhodanica from La Grive L7 (MN7). These taxa display a highly derived lower dentition, with elongated and relatively narrow molars with highly reduced cingula, among other distinctive features (e.g., Andrews et al.,1996; Begun,2002). However, another group composed by the Late Miocene crouzeliines Anapithecus (MN9; Kretzoi,1975; Begun,2002; Nargolwalla et al.,2005), Egarapithecus (Moyà-Solà et al., 2011; MN10) and Laccopithecus (MN11; Wu and Pan,1984)—here included into a distinct tribe, Anapithecini—can be distinguished by the former by their more conservative occlusal morphology (as exemplified among others by the retention of more developed cingula in Egarapithecus and, to a lesser extent, Anapithecus).

Begun (2002) further included into the genus Plesiopliopithecus the mandibular fragment with M/1 and damaged M/2 from Priay (MN9), previously attributed to Pliopithecus priensis by other authors (Welcomme et al.,1991; Andrews et al.,1996). However, while we agree with the crouzeliine affinities of this taxon, in our opinion its closest affinities lie amongst the Anapithecini (see later). Unfortunately, crouzeliins are much more incompletely known than anapithecins (no upper teeth have been thus far recorded in the former), although on the basis of available remains, they seem more derived than anapithecins, which is clearly at odds with the older chronology of the former. The closer similarities between Barberapithecus and Anapithecus amongst the Crouzeliinae suggest that the former must be included into the newly erected subtribe Anapithecini, probably representing its basalmost member. The M/1 of “Pliopithecus” priensis somewhat resembles that of Barberapithecus by the retention of a centrally situated hypoconulid and the fact that the hypoconid and entoconid are aligned with one another. Nevertheless, the crouzeliine from Priay somewhat more closely resembles Anapithecus by the presence of a longer central fovea and a more reduced distal fovea as compared to Barberapithecus. On this basis, we provisionally attribute the remains from Priay to cf. Anapithecus priensis comb. nov., which further agrees with the MN9 age of its type locality. Additional material would be required to confirm whether this species belongs to Anapithecus or to another genus, although the scanty available evidence supports our contention that Barberapithecus, although more primitive, is closely related to Anapithecus and alleged taxa.

A possible alternative phylogenetic interpretation would be considering Barberapithecus as a derived pliopithecinae, evolved from a Pliopithecus-like ancestor that convergently acquired several crouzeliine-like features because of similar trophic adaptations. It should be stressed, however, that the features shared between Barberapithecus and the crouzeliine Anapithecus are not restricted to overall occlusal morphology (cusp shape and crest development), but further include many other independent, derived crouzeline features in several teeth, such as: the long preprotocrista and short hypoparacrista, resulting in a restricted mesial fovea on the upper molars; the distinctive metaconid at the end of the hypoprotocristid on the P/3; the well-developed premetacristid on the P/4; the deep trigonid foveae in the P/4 and lower molars; and the small-sized entoconid on the lower molars. Other crouzeliines, especially Laccopithecus and Egarapithecus, share several (but not all) of these features: thus, Egarapithecus lacks a well-developed premetacristid on the P/4 (resulting in a lingually open trigonid), whereas the presence of a distinct P/3 metaconid is variable in Laccopithecus, and Plesiopliopithecus lacks a distinct P/3 metaconid at the end of the hypoprotocristid. Similarly, previously known crouzeliines share several other features to the exclusion of Barberapithecus. These features, such as the low-crowned upper incisors (known in Anapithecus and Laccopithecus), the reduced cingula on the upper and lower molars, the long and narrow central foveae in the lower molars, and the buccally situated hypoconulid on the M/1 and M/2, are likely to be derived, thus suggesting a basal position for Barberapithecus amongst the Anapithecini. However, given the mosaic distribution of primitive and derived features amongst the Crouzeliinae—which is even more marked after the inclusion of Barberapithecus within the latter—it seems undeniable that crouzeliine-like features must have independently evolved between several taxa to some extent.

The material of the youngest Pliopithecus species, P. canmatensis from several localities of Abocador de Can Mata (Alba et al.,2010), may provide relevant information for investigating the possible phylogenetic link between Barberapithecus and pliopithecines. The material of P. canmatensis, which comes from five different localities that comprise a restricted time span on only 0.2 myrs, show a considerable degree of intraspecific variation in occlusal morphology. Three different morphotypes can be distinguished. Thus, in the oldest locality (ACM/C5-C3), the lower teeth are relatively wider, further displaying more bulbous cusps and blunter crests, and small mesial and distal foveae on the molars, with the M/3 being less elongated than in the remaining sites; in these respects, specimens of P. canmatensis from this locality are more similar to Pliopithecus antiquus from France. The lower teeth from most other localities (ACM/C4-Cb, ACM/C5-C2, and ACM/C4-A1, the latter being the type locality of P. canmatensis) are more elongated and display a greater development of cingulids. Finally, the material from ACM/C5-A8 (intermediate in age between ACM/C5-C2 and ACM/C4-A1) is the most distinctive, with the lower dentition showing more buccolingually compressed cusps and sharper crests, and with the mesial and distal foveae being larger than in the remaining localities; moreover, the lower molars display a short posthypocristid, and the M/2 and M/3 central foveae are short and quadrangular; in all these respects, the material from ACM/C5-A8 somewhat resembles the occlusal morphology of Barberapithecus, although crouzeliine features are much more strongly developed in the latter. It must be stressed that within the known populations of P. canmatensis there is no temporal trend from a more primitive (P. antiquus-like) occlusal morphology towards a more derived morphotype (reminiscent of Barberapithecus): thus, the latter is only present in a locality (ACM/C5-A8), which is somewhat older than the type locality (C4-A1), where the typical morphotype of P. canmatensis is recorded. Nevertheless, the morphotype recorded at ACM/C5-A8 leaves open the possibility that a crouzeliine like Barberapithecus might have evolved from a pliopithecine ancestor similar to Pliopithecus canmatensis.

Even if Barberapithecus had evolved from a Pliopithecus ancestor, this would not necessarily imply that the former must be excluded from the Crouzeliinae. Indeed, Barberapithecus may be considered a stem anapithecin reflecting the ancestral occlusal morphology of this group. In this view, the two possible phylogenetic interpretations outlined above for Barberapithecus (an anapithecin closely related to Anapithecus vs. a taxon derived from a Pliopithecus-like ancestor that converged with true crouzeliines) are not necessarily mutually exclusive, as long as it is recognized that, from a taxonomic viewpoint, only crouzeliins (i.e., Plesiopliopithecus) can be regarded as undoubted true crouzeliines (the attribution of anapithecins to the Crouzeliinae depending on their supposed closest affinities with Plesiopliopithecus than with Pliopithecus). It is certainly conceivable that crouzeliins and anapithecins might not share a single common ancestor to the exclusion of pliopithecins, with the former having evolved from a dionysopithecine or pliopithecine ancestor in either Europe or Asia by the latest Early Miocene, and the latter having evolved from a derived Pliopithecus species by the latest Middle Miocene in Europe (as suggested by the similarities simultaneously shown by Barberapithecus with P. canmatensis, on the one hand, and Anapithecus and other anapithecins, on the other). This hypothesis would account for the old chronology of Plesiopliopithecus as compared to other crouzeliines, which contrasts with the apparently much derived occlusal morphology of the former. From a taxonomic viewpoint, this phylogenetic hypothesis would not only render the Pliopithecinae as paraphyletic, but would imply that the Crouzeliinae as currently conceived are polyphyletic. The limited dental evidence currently available for the several species of Plesiopliopithecus, together with the mosaic distribution of dental features among the Pliopithecidae and, in particular, the Crouzeliinae, seriously hampers conducting a reliable phylogenetic assessment of Barberapithecus. An undescribed new pliopithecid genus from the late early Miocene of Fanchang (China), preliminarily interpreted as a stem crouzeliine (Harrison and Jin,2009), might help to clarify this issue in the future. Until the latter taxon is described or a cladistic analysis of dental morphology amongst the Pliopithecidae is performed (which is outside the scope of this article), we currently favor the view that the similarities between Barberapithecus and Anapithecus are homologous and indicate that the former is a stem anapithecin, whereas Plesiopliopithecus is a more derived taxon that might have independently evolved from a different pliopithecid ancestor. Given the profound taxonomic implications of the latter hypothesis, however, we provisionally prefer to take the conservative view of preserving the integrity of the Crouzeliinae as one of the three pliopithecid subfamilies (e.g., Moyà-Solà et al.,2001; Alba et al.,2010), merely separating the anapithecins from the crouzeliins at the tribe rank.

SUMMARY AND CONCLUSIONS

A new genus and species of fossil catarrhine primate, Barberapithecus huerzeleri (Pliopithecidae: Crouzeliinae: Anapithecini) is described based on dental material from Castell de Barberà. On the basis of the lack of Cricetulodon and the presence of Hippotherium at least from a layer stratigraphically a few meters above the main fossiliferous layer (which yielded the pliopithecid remains), the newly erected taxon corresponds to the latest Aragonian (Middle Miocene) or earliest Vallesian (Late Miocene), with an estimated age of about 11.2–10.5 Ma. The holotype of Barberapithecus includes 15 teeth from a single female individual—representing all the permanent teeth except the I2/, the I/1, the I/2, the P3/, and the P4/—as well as by an isolated P/3 (paratype) and a male C1/ from the same locality (tentatively referred to the hypodigm of the new taxon, but alternatively interpreted by some authors as a female hominoid canine). Barberapithecus huerzeleri shows a mosaic of pliopithecine and crouzeliine features, combined with several autapomorphic features, which justify the erection of a new genus. The latter features include, among others, the very buccolingually compressed and mesiodistally oriented main cusp of the C1/ and C/1; the large P/4 trigonid subequal in size to the talonid; the lower molars with quadrangular central foveae and mesially situated entoconid; and the large distal foveae in the M/1 and especially the M/2. Despite some dental similarities with Pliopithecus, Barberapithecus displays a set of derived crouzeliine synapomorphies, such as: buccolingually compressed cusps; well-developed and relatively sharp crests; large and well-defined occlusal foveae; peripheralized cusps; upper molars with somewhat distally situated protocone, long preprotocrista, short hypoparacrista, and short distal fovea; distinct P/3 metaconid; well-developed P/4 premetacristid; and lower molars with reduced entoconid. Based on these derived features, we formally attribute Barberapithecus to the subfamily Crouzeliinae. Within this group, Barberapithecus most closely resembles the Late Miocene genera Anapithecus and, to a lesser extent, the younger and more derived Laccopithecus and Egarapithecus. On this basis, we include Barberapithecus, together with the latter taxa, into a newly erected tribe Anapithecini, which differs from the monotypic Crouzeliini (including only Plesiopliopithecus) by the retention of several primitive features.

The retention of several occlusal features resembling Pliopithecus species, as shown by the comparisons with P. canmatensis from the late Middle Miocene (ca. 11.8–11.7 Ma) of the Vallès-Penedès Basin, as well as the dental proportions of Barberapithecus (intermediate between dionysopithecines/pliopithecines and crouzeliins), suggest that Barberapithecus might have evolved from a Pliopithecus-like ancestor such as the more derived populations of P. canmatensis, further being the basalmost member of the Anapithecini. Given the distribution of primitive and derived features amongst the Crouzeliinae, some degree of homoplasy seems unavoidable regarding the evolution of crouzeliine purported synapomorphies. This, coupled with the much more derive dental occlusal pattern of Plesiopliopithecus in spite of their older chronology as compared to Barberapithecus and other anapithecins, suggests that as currently conceived the Crouzeliinae might be polyphyletic, with crouzeliins and anapithecins having independently evolved from different pliopithecid ancestors in a different time/place. Unfortunately, the currently available material of Plesiopliopithecus is too scarce to test confidently this hypothesis, although the description of additional purported crouzeliine material from the late Early Miocene of China might help to clarify this issue in the future. For the moment being, we provisionally retain the taxonomic integrity of the Crouzeliinae, although noting that they probably constitute an unnatural assemblage, with anapithecins having probably independently evolved from a derived Pliopithecus species in Europe during the latest Middle Miocene.

Acknowledgements

The authors thank Marta Palmero for the splendid drawings, Daniel De Miguel for assistance when making the ESEM micrographs, and Eric Delson and David Begun for cast exchange. They further acknowledge the comments and suggestions by two anonymous reviewers, who helped them to improve a previous version of this article.